Failure of pulmonary gas exchange is most commonly due to alveolar flooding diseases as in pulmonary edema, or to acute airflow obstruction, as in status asthmaticus. In both conditions, conventional mechanical ventilation (CMV) is associated with significant complications, especially barotrauma and reduction of cardiac output. One long-term goal of this study is to explore innovative settings of CMV and its combination with continuous flow ventilation (CFV); another is to gain greater understanding of the unusual gas mixing between inspired and alveolar gas during CFV, and how this process might be linked to bronchial aerodynamics and to the site and nature of the V/Q variance within normal lungs. Recent studies of CFV demonstrate unexpected hyperinflation of the lungs, associated with lobar inhomogeneity of inflation, arterial hypoxemia, and 4-5 fold increases in airways resistance compared to CMV. In a first series of studies, we propose: a) to describe and explain the effects of gas flow rate and physical properties on airway dynamics during CFV in a bronchial model, and to extend these measurements to the anesthetized dog, utilizing retrograde catheter techniques; b) to relate the observed airways dynamics to lung hyperinflation, and to explore the use of low density gas (heliox) and position of the continuous flow jets within the airway to minimize the lobar inhomogeneity; c) to distinguish whether incomplete gas phase diffusion or increased V/Q variance is the cause of arterial hypoxemia utilizing the multiple inert gas technique applied to the whole lung and to the hyperinflated lower lobes during CFV; d) to evaluate newer combinations of CFV with small tidal volume excursions of CMV to optimize the best advantages of both forms of mechanical ventilation. In a second series of studies we intend to use CFV as an experimental tool to explore animal models of respiratory failure. In canine pulmonary edema: a) we will use the hysteresis of the lung pressure-volume relationship to seek the optimal alveolar pressure (Palv) at which shunt is minimized; b) at constant Palv, we will relate changes in lung volume and shunt to deduce mechanisms of edema redistribution between interstitial and alveolar spaces; c) comparison of these results with CMV will assess the role of cyclic ventilation on edema distribution and accumulation. In models of canine and porcine airflow obstruction, we will measure the effects of bronchospasm and gas trapping on multiple inert gas exchange during CMV and CFV, and test the effect of CFV to minimize gas trapping. One long-term objective of these studies is to provide greater understanding concerning the interrelationships among alveolar pressure, lung volume, pulmonary gas exchange, and the inciting disease (edema or bronchospasm). The elucidated mechanisms suggest several proposed extensions of alternative modes of ventilation to the clinical conditions being modelled.